From fiction to friction: abusing autonomous mobile robots

This dissertation investigates the emerging phenomenon of human violence toward autonomous mobile robots (AMRs), focusing on a food delivery robot as the empirical context. Drawing on theories of dehumanization, four experiments examined how individuals attribute cognitive humanness to robots, whether violence toward AMRs stems from instrumental versus moral motives, and how framing (humanized vs. dehumanized) and job displacement narratives shape willingness to aggress. The Humanity Inventory was used to assess dispositional tendencies to dehumanize, and a new measure, the Humanness Attribution Scale (HAS), was developed to capture how people ascribe Uniquely Human (UH) and Human Nature (HN) traits to robots. Across studies, participants were more likely to aggress for instrumental gain than moral justification; making AMRs appear more humanlike did not reduce aggression, and job displacement framing did not influence robot abuse. Dehumanization consistently emerged as a mechanism normalizing harm, though the Humanity Inventory showed inconsistent evidence for stable dispositional traits underlying dehumanization. The HAS demonstrated meaningful variability in how participants attributed humanness to AMRs, supporting its utility in HRI research. Findings suggest that anthropomorphic design alone does not mitigate aggression toward robots. Instead, broader situational, cultural, and cognitive factors strongly shape human responses. This work highlights the ethical, social, and psychological complexities of robot abuse and calls for interdisciplinary solutions integrating behavioral science, technology design, and policy.